Extracellular Visfatin-Promoted Malignant Behavior in ... · Biology of Human Tumors Extracellular...

Biology of Human Tumors

Extracellular Visfatin-Promoted MalignantBehavior in Breast Cancer Is Mediated Throughc-Abl and STAT3 ActivationAmos C. Hung1, Steven Lo2, Ming-Feng Hou3, Yi-Chen Lee1,4, Chun-Hao Tsai1,5,Yuan-Yin Chen1,Wangta Liu6, Yu-Han Su1, Yi-Hsuan Lo1, Chie-Hong Wang1,Shiou-Chen Wu7, Ya-Ching Hsieh8, Stephen Chu-Sung Hu9,10, Ming-Hong Tai11,Yun-Ming Wang7, and Shyng-Shiou F. Yuan1,5,12,13

Abstract

Purpose: Visfatin is an adipocytokine involved in cellularmetabolism, inflammation, and cancer. This study investigatedthe roles of extracellular visfatin in breast cancer, and exploredunderlying mechanisms in clinical and experimental settings.

Experimental Design: Associations of serum visfatin withclinicopathologic characteristics and patient survival wereassessed with Cox regression models and Kaplan–Meier anal-yses. Effects of extracellular visfatin on cultured breast cancercells were examined, followed by in vivo investigation of tumorgrowth and metastasis in xenograft animal models. Imatiniband Stattic were used to inhibit c-Abl and STAT3 activation,respectively.

Results: Breast cancer patients with high serum visfatin levelswere associated with advanced tumor stage, increased tumor sizeand lymph node metastasis, and poor survival. Elevated phos-phorylation of c-Abl and STAT3 in breast tumor tissues were

correlated with high serum visfatin levels in patients. Visfatin-promoted in vitro cell viability and metastatic capability weresuppressed by imatinib (c-Abl inhibitor) and Stattic (STAT3inhibitor). Increased in vivo cell invasiveness was observed inzebrafish xenografted with visfatin-pretreated breast cancer cells.Tumor growth and lung metastasis occurred in visfatin-adminis-tered mice xenografted with breast cancer cells. Tail vein–injectedmice with visfatin-pretreated breast cancer cells showed increasedlung metastasis, which was suppressed by imatinib.

Conclusions: Serum visfatin levels in breast cancer patientsreveal potential prognostic values, and our findings that vis-fatin promoted breast cancer through activation of c-Abl andSTAT3 may provide an important molecular basis for futuredesign of targeted therapies that take into account differentserum visfatin levels in breast cancer. Clin Cancer Res; 22(17);4478–90. �2016 AACR.

IntroductionBreast cancer is one of the most common malignancies in

women worldwide, with multiple risk factors identified, includ-ing differential expression of adipocytokines (1, 2). Visfatin, alsoknown as nicotinamide phosphoribosyltransferase (NAMPT) orpre–B-cell colony-enhancing factor (PBEF), is a 52 kDa adipocy-tokine discovered both intracellularly and extracellularly (3).Intracellular visfatin functions as a rate-limiting enzyme in thebiosynthesis of nicotinamide adenine dinucleotide (NAD).Whenreleased outside of cells, visfatin exhibits dual roles in enzyme-

like activity on extracellular NAD formation as well as cytokine-like activity through a putative receptor–mediated pathway (4).Both intracellular and extracellular visfatin have been shown tobe involved in the tumor development (5).

We previously reported that high intracellular expression ofvisfatin in breast tumor tissues was associated with poor patientsurvival (6). In recent years, the correlation of serum visfatinlevel with breast cancer behavior has started to emerge fromdifferent cohort studies (7–11). However, the mechanismslinking biologic functions of extracellular visfatin to its clinical

1Translational Research Center, Kaohsiung Medical University Hos-pital, Kaohsiung Medical University, Kaohsiung, Taiwan. 2Cannies-burn Plastic Surgery Unit, Royal Infirmary, Glasgow, Scotland, Unit-ed Kingdom. 3Cancer Center, Kaohsiung Medical University Hospi-tal, Kaohsiung Medical University, Kaohsiung, Taiwan. 4Departmentof Anatomy, College of Medicine, Kaohsiung Medical University,Kaohsiung, Taiwan. 5Graduate Institute of Medicine, College ofMedicine, Kaohsiung Medical University, Kaohsiung, Taiwan.6Department of Biotechnology, College of Life Science, KaohsiungMedical University, Kaohsiung, Taiwan. 7Department of BiologicalScience and Technology and Institute of Molecular Medicine andBioengineering, National Chiao Tung University, Hsinchu, Taiwan.8Department of Medical Research, E-Da Hospital, I-Shou University,Kaohsiung, Taiwan. 9Department of Dermatology, KaohsiungMedical University Hospital, Kaohsiung, Taiwan. 10Department ofDermatology, College of Medicine, Kaohsiung Medical University,Kaohsiung, Taiwan. 11Institute of Biomedical Sciences, National SunYat-Sen University, Kaohsiung, Taiwan. 12Department of Obstetricsand Gynecology, Kaohsiung Medical University Hospital, Kaohsiung

Medical University, Kaohsiung, Taiwan. 13Department of MedicalResearch, Kaohsiung Medical University Hospital, Kaohsiung Med-ical University, Kaohsiung, Taiwan.

Note: Supplementary data for this article are available at Clinical CancerResearch Online (http://clincancerres.aacrjournals.org/).

Corresponding Authors: Shyng-Shiou F. Yuan, Translational Research Center,Kaohsiung Medical University Hospital, Kaohsiung Medical University, 100Zihyou 1st Road, San-Ming District, Kaohsiung 807, Taiwan. Phone: 886-7-312-1101, ext. 2557; Fax: 886-7-311-2493; E-mail: [email protected]; andYun-MingWang, Department ofBiological Science andTechnology and Instituteof Molecular Medicine and Bioengineering, National Chiao Tung University, 75Bo-Ai Street, Hsinchu 300, Taiwan. Phone: 886-3-571-2121, ext. 56972; E-mail:[email protected]

doi: 10.1158/1078-0432.CCR-15-2704

�2016 American Association for Cancer Research.

ClinicalCancerResearch

Clin Cancer Res; 22(17) September 1, 20164478

on April 25, 2021. © 2016 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst April 1, 2016; DOI: 10.1158/1078-0432.CCR-15-2704

http://clincancerres.aacrjournals.org/

significance in breast cancer remain unknown. In this study, weinvestigated the potential clinical value of circulating visfatin inbreast cancer. In addition, we evaluated the functional roles ofextracellular visfatin in breast tumor growth and metastaticcapability using both in vitro and in vivo models. The molecularmechanisms underlying these observations were also explored.

Materials and MethodsPatient samples

Two hundred and fifty-eight female patients with patholog-ically confirmed invasive ductal carcinoma of the breast wereincluded in this study. Sera andbreast tumor tissueswere collectedfrom surgically treated patients at Kaohsiung Medical UniversityHospital (KMUH), Taiwan, from 2001 to 2011. After surgery, thepatients were administered adjuvant radiotherapy, chemothera-py, or hormone therapy based on the clinical practice guidelinesof breast cancer (12). Hormone therapy was administered topatients whose tumor sections were estrogen receptor (ER) pos-itive. This study was approved by the Institutional Review Boardof KMUH, and informed consent was obtained from each patient.Histologic type and grading of the primary tumor were assessedaccording to the World Health Organization (WHO) classifica-tion (13) and the modified Bloom–Richardson grading scheme(14), respectively, and staging was evaluated according to theAmerican Joint Committee on Cancer (AJCC) TNM staging sys-tem (15). The status of ER, progesterone receptor (PR), and HER2was determined by immunohistochemical analysis.

Enzyme immunoassaySerum visfatin levels for breast cancer patients (n ¼ 258) and

age-matched normal female participants (n ¼ 100) were mea-sured in duplicate by a human visfatin-specific enzyme immu-noassay kit (Phoenix Pharmaceuticals) according to the manu-facturer's instructions.

Tissue microarrayTumor tissue samples of the patients were obtained from for-

malin-fixed and paraffin-embedded tissue blocks for the construc-

tion of tissue microarray by an Alphelys BoostArrayer device(Plaisir) as described previously (6). Five-micrometer sections fromthe tissuemicroarray were obtained using amicrotome and immu-nohistochemically stained for determinationof protein expression.

ImmunohistochemistryTheprocedureof immunohistochemical stainingwasdescribed

previously (6). The primary antibodies used for immunohis-tochemistry in this study included the mouse monoclonal anti-ER, anti-PR, and anti-HER2 antibodies from DAKO (Glostrup),rabbit polyclonal anti–phospho-c-Abl (Y393/412) and anti-vimentin antibodies from GeneTex (Irvine), rabbit monoclonalanti–phospho-STAT3 (Y705) antibody from Cell Signaling Tech-nology, and mouse monoclonal anti-cytokeratin 18 antibodyfrom Leica. Images of immunohistochemically stained sectionswere captured by a Nikon Eclipse E600 microscope. For thescoring of protein expression in breast tumor tissues, the stainingof phospho-c-Abl (Y393/412) and phospho-STAT3 (Y705) wasstratified into quartiles (0, undetectable; 1, low; 2, intermediate; 3,high)on thebasis of intensity as describedpreviously (16). For theanalysis of tissues from mice, the staining of phospho-c-Abl(Y393/412), phospho-STAT3 (Y705), cytokeratin 18, and vimen-tin was scored by the method of histochemical score (H-score),which was calculated as the product of percentage of stained cellsand intensity of staining (17).

Cell cultureHuman breast cancer cell lines MDA-MB-231, MCF7, and

T-47D were purchased from the Bioresource Collection andResearch Center of Taiwan with authentication for genotypes andphenotypes of the cells. All three cell lines were cultured inDMEM(Life Technologies) supplemented with 10% FBS (BiologicalIndustries), 100 U/mL penicillin (Biological Industries), 100mg/mL streptomycin (Biological Industries), and 2.5 mg/mLamphotericin B (Biological Industries) at 37�C in a humidified5% CO2 incubator. The MDA-MB-231 cells stably expressingluciferase (MDA-MB-231-Luc) were kindly provided by Prof.Wen-Chun Hung, National Health Research Institutes, Taiwan.

XTT cell viability assayViable cells were determined by the tetrazolium-based XTT

colorimetric assay (Sigma-Aldrich) according to the manufac-turer's instructions. MDA-MB-231, MCF7, and T-47D cells weretreated with different concentrations of visfatin (0–100 ng/mL)for 72 hours before the XTT assay. For experiments involvingimatinib (c-Abl inhibitor; ref. 18) and Stattic (STAT3 inhibitor;ref. 19), MDA-MB-231 cells were pretreated with imatinib (0 or10 mmol/L) and Stattic (0 or 1 mmol/L) for 1 hour, followed byvisfatin (0–100 ng/mL) treatment for 72 hours.

ImmunoblottingTheprocedures of protein extraction and immunoblottingwere

performed as described previously (20).MDA-MB-231 andMCF7cells were treated with different concentrations of visfatin (0–100ng/mL) for 2 hours before the immunoblotting analysis. Theprimary antibodies used for immunoblotting in this study includ-ed the rabbit polyclonal anti–phospho-c-Abl (Y393/412), anti–c-Abl, anti-STAT3, anti-JAK2 antibodies and rabbit monoclonalanti–phospho-JAK2 (Y1007/1008) antibody from GeneTex, rab-bit monoclonal anti–phospho-STAT3 (Y705) and anti-HER2antibodies from Cell Signaling Technology, rabbit polyclonal

Translational Relevance

Visfatin is known to have a number of important biologicactivities, including nicotinamide adenine dinucleotide(NAD) biosynthesis, cellular metabolism, inflammatoryresponse, and cancer progression. Although previous studieshave suggested a role for serum visfatin as a biomarker inbreast cancer, the molecular mechanisms linking biologicfunctions of extracellular visfatin to its clinical significanceremain unidentified. In this study, we report that activation ofc-Abl andSTAT3 is involved in extracellular visfatin–promotedbreast cancer via clinical and experimental investigations.Moreover, the results of our in vitro and in vivo studies showthat imatinib, a tyrosine kinase inhibitor currently used forclinical treatment of multiple cancers but not for breast canceryet, inhibits extracellular visfatin-induced malignant beha-viors. This further raises a possibility that future design ofclinical trials for evaluation of therapeutic interventions, suchas that for imatinib, may benefit from stratification of breastcancer patients according to different serum visfatin levels.

Role of Extracellular Visfatin in Breast Cancer

www.aacrjournals.org Clin Cancer Res; 22(17) September 1, 2016 4479




anti–phospho-STAT3 (S727) antibody from Santa Cruz Biotech-nology, and mouse monoclonal anti–b-actin antibody fromSigma-Aldrich. The chemiluminescent signal was captured byChemiDoc XRSþ System (Bio-Rad Laboratories) and quantifiedwith Image Laboratory software (Bio-Rad Laboratories).

In vitro cell migration and invasion assaysIn vitro cell migration was determined using Transwell inserts

with 8-mm pores (Corning). MDA-MB-231 cells were pretreatedwith imatinib (0 or 10 mmol/L) and Stattic (0 or 1 mmol/L) for1 hour, followed by visfatin (0 or 100 ng/mL) pretreatment for

Months

0 24 48 72 96 120

Dis

ease

-free

sur

viva

l

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Visfatin low/ER+Visfatin low/ER−Visfatin high/ER+Visfatin high/ER−

P < 0.001

No. of patients at risk56 54 38 24 14 226 24 17 11 9 0

113 106 65 23 20 163 52 25 11 6 0

Variable HR (95% CI)Visfatin low/ER+ 1.00 (referent)Visfatin low/ER− 2.55 (0.51−12.65)Visfatin high/ER+ 4.60 (1.37−15.38)Visfatin high/ER− 5.95 (1.73−20.42)

Months0 24 48 72 96 120

Dis

ease

-free

sur

viva

l

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Visfatin lowVisfatin high

Variable HR (95% CI)Visfatin low 1.00 (referent)Visfatin high 3.54 (1.48−8.47) P < 0.001

82 79 56 36 24 3176 159 66 35 27 2

No. of patients at risk

B

Months0 24 48 72 96 120

Ove

rall

surv

ival

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Visfatin lowVisfatin high

No. of patients at risk82 80 57 36 24 3

176 166 75 42 31 2

Variable HR (95% CI)Visfatin low 1.00 (referent)Visfatin high 4.18 (1.45−12.03) P < 0.001

C

Months0 24 48 72 96 120

Ove

rall

surv

ival

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Visfatin low/ER+Visfatin low/ER−Visfatin high/ER+Visfatin high/ER−

P = 0.001

No. of patients at risk56 54 39 24 14 226 24 17 11 9 0

113 109 46 27 23 163 56 28 14 7 0

Variable HR (95% CI)Visfatin low /ER+ 1.00 (referent)Visfatin low/ER− 7.89 (0.82−76.01)Visfatin high/ER+ 10.61 (1.41−79.98)Visfatin high/ER− 14.49 (1.90−110.61)

ED

A

Control

Breast cancer

Ser

um v

isfa

tin(n

g/m

L)

0

20

40

60

80

100P < 0.001

Control

Breast cancer

BM

I(k

g/m

2 )

10

15

20

25

30

35

40P < 0.001

Control

Breast cancer

Ser

um v

isfa

tin/B

MI

([ng/

mL]

/[kg/

m2 ])

0

1

2

3

4

5P < 0.001

Figure 1.Comparison of serum visfatin levels between breast cancer patients and normal participants, and association of serum visfatin levels with patient survival inbreast cancer. A, serum visfatin (left) was measured by enzyme immunoassay. BMI (middle) was measured and used for normalization of serum visfatin(serum visfatin/BMI; right). Averaged serum visfatin in normal female participants and breast cancer patients were 32.20 � 17.42 and 40.87 � 13.86 ng/mL,respectively. Averaged BMI in normal female participants and breast cancer patients were 21.97 � 3.20 and 23.42 � 3.61 kg/m2, respectively. Averaged serumvisfatin/BMI in normal female participants and breast cancer patients were 1.47 � 0.78 and 1.77 � 0.65 (ng/mL)/(kg/m2), respectively. The data were presentedby box plots, where the lower boundary and top boundary of the box indicated the 25th percentile and 75th percentile, respectively. The line within the boxmarked the median, and the bars below and above the box indicated the 10th and 90th percentiles. P values were determined by two-sided Student t test.Control indicated normal female participants (n¼ 100). Breast cancer indicated breast cancer patients (n ¼ 258). B to E, patient survival was analyzed by Kaplan–Meier curves. The HRs and 95% confidence intervals (CI) were calculated using a Cox regression model. B, disease-free survival was analyzed according to serumvisfatin levels in breast cancer patients. C, OS was analyzed according to serum visfatin levels in breast cancer patients. D, disease-free survival was analyzedaccording to combined serum visfatin levels and ER status in breast cancer patients. E, OS was analyzed according to combined serum visfatin levels andER status in breast cancer patients. P values were determined by two-sided log-rank test.

Hung et al.

Clin Cancer Res; 22(17) September 1, 2016 Clinical Cancer Research4480




24 hours. The cells were then trypsinized and resuspended inserum-free DMEM, and added to the Transwell inserts (3 � 104

cells/insert) assembled on a 24-well plate, in which the wellscontained DMEM with 10% FBS. After 24 hours of incubation,cells remaining on the upper surface of the membrane insidethe inserts were removed with cotton swabs, whereas cells thatmigrated to the underside of the membrane were stained with0.5 g/L crystal violet (Sigma-Aldrich). The cells on the under-side of the membrane were imaged by an Olympus SZX10microscope, and quantification of the total area that the cellsoccupied was processed with the NIS-Elements software (Lab-oratory Imaging).

In vitro cell invasion was determined using Transwell insertswith 8-mm pores coated with Matrigel (Corning), and the proce-dure was performed as described previously (21), with the sameprotocol of cell treatment as described above.

Gelatin zymographyThe procedure of gelatin zymography for secreted matrix

metalloproteinases-2/9 was described previously (21). MDA-MB-231 cells were pretreated with imatinib (0 or 10 mmol/L)and Stattic (0 or 1 mmol/L) for 1 hour, followed by visfatin (0 or100 ng/mL) treatment for 48 hours. The supernatants from eachexperimental condition were collected for gelatin zymographicanalysis.

In vivo cell invasiveness in zebrafishThe experiments using zebrafish in this studywere approved by

the Institutional Animal Care and Use Committee of KaohsiungMedical University. Embryos of zebrafish (strain fli1:EGFP fromthe Taiwan Zebrafish Core Facility; http://tzcf-hdmrc.org/) weregenerated by natural pairwise mating (22), and the procedure ofxenograft was based on a previous report (23). In brief, MDA-MB-231 cells pretreated with visfatin (0 or 100 ng/mL) for 24 hourswere labeled with fluorescent probe DiI (Life Technologies), andimplanted into the perivitelline cavity of 2-day-old zebrafishembryos through microinjection. After confirmation of the local-ized DiI-labeled cell mass at the injection site, the zebrafish weretransferred to fresh water and maintained at 32.5�C for 48 hours,and cell invasion was determined by visualizing dissemination ofthe DiI-labeled cells from the injection site under a Nikon EclipseTi-S microscope.

Soft agar colony formation assayThe procedure of anchorage-independent colony formation

in soft agar was described previously (21). MDA-MB-231 andMCF7 cells grown in the soft agar were treated with visfatin(0–100 ng/mL; three times a week) for 3 weeks.

In vivo orthotopic tumor growth and distant metastasis inmice

Six-week-old female immunodeficient Foxn1nu/Foxn1nu mice(nudemice) were obtained from the National Laboratory AnimalCenter of Taiwan (http://www.nlac.org.tw/). All of the miceexperiments in this study were approved by the InstitutionalAnimal Care and Use Committee of Kaohsiung Medical Univer-sity.MDA-MB-231 cells (2.5�106 cells) resuspended in100mLofnormal saline were injected into the fourth mammary fat pads ofmice. One week after implantation, the mice were randomlydivided into two groups (n ¼ 10/group), followed by intraper-itoneal administration of visfatin (0.5 mg/kg) or normal saline

three times a week. The tumor size was measured weekly andcalculated by the formula of (width2 � length)/2. After 8 weeks,all mice were sacrificed and the orthotopic tumors were collectedfor tumor weight measurement and immunohistochemical anal-ysis. The lung tissues were examined for distant metastasis, whichwas determined by immunohistochemical analysis for the expres-sionof cytokeratin 18andvimentin, two tumormarkers expressedin MDA-MB-231 cells (24).

In vivo bioluminescent imaging of metastasis in miceThe luciferase-expressing MDA-MB-231 (MDA-MB-231-Luc)

cells were pretreated with imatinib (0 or 10 mmol/L) for 1 hour,followed by visfatin (0 or 100 ng/mL) pretreatment for 24 hoursbefore intravenous injection of the cells (1 � 106 cells in 100 mLof normal saline) into the lateral tail vein of nude mice (n ¼ 3–4/group). The mice injected with the cells pretreated withimatinib alone, or the mice injected with the cells pretreatedwith imatinib followed by visfatin pretreatment, were intra-peritoneally administered imatinib (50 mg/kg daily; ref. 25).On day 7 after tail vein injection of the cells, the mice wereanalyzed for the presence of bioluminescent signals by Xeno-gen IVIS Spectrum in vivo imaging system (Caliper LifeSciences). The mice were anesthetized with isofurane (Baxter)

Table 1. Clinicopathologic characteristics of breast cancer patients and theassociation with serum visfatin

Serum visfatina

Low HighVariable Patient, n (%) n (%) n (%) Pd

Total no. 258 (100) 82 (31.8) 176 (68.2) –

Stageb

I 92 (35.7) 34 (41.5) 58 (33.0) <0.001II 118 (45.7) 44 (53.7) 74 (42.0)III and IV 48 (18.6) 4 (4.9) 44 (25.0)

Gradec

1 34 (13.2) 13 (15.9) 21 (11.9) 0.6872 156 (60.5) 48 (58.5) 108 (61.4)3 68 (26.3) 21 (25.6) 47 (26.7)

Age, y�50 138 (53.5) 48 (58.5) 90 (51.1) 0.267>50 120 (46.5) 34 (41.5) 86 (48.9)

BMI (kg/m2)<24 153 (59.3) 57 (69.5) 96 (54.5) 0.023�24 105 (40.7) 25 (30.5) 80 (45.5)

Tumor size (cm)<2 103 (39.9) 43 (52.4) 60 (34.1) <0.0012–5 119 (46.1) 36 (43.9) 83 (47.2)>5 36 (14.0) 3 (3.7) 33 (18.8)

LN metastasis0–1 172 (66.7) 65 (79.3) 107 (60.8) 0.003�2 86 (33.3) 17 (20.7) 69 (39.2)

ER statusNegative 89 (34.5) 26 (31.7) 63 (35.8) 0.520Positive 169 (65.5) 56 (68.3) 113 (64.2)

PR statusNegative 118 (45.7) 32 (39.0) 86 (48.9) 0.140Positive 140 (54.3) 50 (61.0) 90 (51.1)

HER2 statusNegative 167 (64.7) 62 (75.6) 105 (59.7) 0.013Positive 91 (35.3) 20 (24.4) 71 (40.3)

Abbreviations: LN, lymph node; –, not applicable.aThe cutoff value of low (<33.75 ng/mL) and high (�33.75 ng/mL) serumvisfatinwas determined by the ROC curve.bStaging was based on the AJCC TNM staging system (15).cGrading was based on the modified Bloom-Richardson grading scheme (14).dP values were calculated by two-sided c2 test.






using a XGI-8 Gas Anesthesia System (Caliper Life Sciences)followed by intraperitoneal injection of D-luciferin (150 mg/kg;PerkinElmer) for the detection of luciferase expression. The opti-cal images were acquired and analyzed by Xenogen Living Imagesoftware (Caliper Life Sciences).

Statistical analysisAll statistical analyses were performed using the SPSS 14.0

statistical package (SPSS). The cutoff point for high and low serumvisfatin level was determined by the receiver operating character-istic (ROC) curve. The associations between visfatin levels andclinicopathologic characteristics were analyzed by the c2 test.Survival curves were generated using the Kaplan–Meier estimatesand analyzed by the log-rank test. Univariate and multivariableanalyses were performed using the Cox proportional hazardsregression models for evaluation of the association betweensurvival and clinicopathologic characteristics. The associationbetween visfatin levels and disease recurrence was analyzed bythe Fisher's exact test or the c2 test. The Student t test was used forcomparison between two groups. All results were consideredstatistically significant if P was less than 0.05 calculated with anappropriate two-sided statistical test.

ResultsClinical association of serum visfatin level with breast cancerprogression

The levels of serum visfatin in breast cancer patients (n ¼258) and normal female participants as controls (n¼ 100) wereassessed by enzyme immunoassay. As shown in Fig. 1A, theserum visfatin levels were significantly increased in breastcancer patients (P < 0.001). In addition, the serum visfatinnormalized for individual body mass index (BMI) was signi-ficantly higher in breast cancer patients (P < 0.001), despitethat the significantly higher BMI was observed in the patients(P < 0.001; Fig. 1A).

The association of patient survival with serum visfatin levelsand their correlation with ER status were analyzed by Kaplan–Meier survival curves, for a follow-up period up to 120 months(median ¼ 48 months). The patients with high levels ofserum visfatin had significantly poorer disease-free survival(P < 0.001; Fig. 1B) and overall survival (OS; P < 0.001; Fig.1C). We further examined the combined association of serumvisfatin levels and ER status with patient survival. The patientswith both high serum visfatin and ER-negative (ER�) statusrevealed the worst disease-free survival (P < 0.001; Fig. 1D).

Table 2. Association between clinicopathologic characteristics of breast cancer patients and OS

Univariatea Multivariablea

Variable HR (95% CI) Pd HR (95% CI) Pe

Gradeb

1 1.00 (referent) – –

2 1.56 (0.53–4.55) 0.4193 2.47 (0.81–7.53) 0.112

Age, y�50 1.00 (referent) – –

>50 1.49 (0.81–2.73) 0.197BMI (kg/m2)<24 1.00 (referent) – –

�24 1.14 (0.60–2.14) 0.690Tumor size (cm)<2 1.00 (referent) 1.00 (referent)2–5 2.47 (1.05–5.81) 0.039 1.68 (0.70–4.05) 0.245>5 4.36 (1.69–11.27) 0.002 2.13 (0.79–5.74) 0.136

LN metastasis0–1 1.00 (referent) 1.00 (referent)�2 3.34 (1.75–6.37) <0.001 2.43 (1.25–4.72) 0.009

HER2 statusNegative 1.00 (referent) – –

Positive 1.14 (0.60–2.17) 0.699RadiotherapyNo 1.00 (referent) – –

Yes 1.11 (0.60–2.03) 0.744ChemotherapyNo 1.00 (referent) – –

Yes 0.89 (0.41–1.93) 0.774Hormone therapyNo 1.00 (referent) 1.00 (referent)Yes 0.36 (0.20–0.66) 0.001 0.46 (0.24–0.87) 0.017

Serum visfatinc

Low 1.00 (referent) 1.00 (referent)High 5.67 (2.01–15.98) 0.001 3.55 (1.22–10.36) 0.020

Abbreviations: LN, lymph node; –, not applicable.aUnivariate and multivariable analyses were performed by Cox regression models. Variables with P values larger than 0.10 in the univariate analysis were excludedfrom multivariable analysis.bGrading was based on the modified Bloom–Richardson grading scheme (14).cLow and high serum visfatin were determined as described in Table 1.dTwo-sided P values were calculated by a univariate Cox proportional hazards regression model.eTwo-sided P values were calculated by a multivariable Cox proportional hazards regression model.

Hung et al.





Similar results for the combined association of serumvisfatin levels and ER status with OS were observed (P ¼0.001; Fig. 1E).

We also analyzed the association of serum visfatin levels withclinicopathologic characteristics in breast cancer patients. Highserum visfatin levels were significantly associated with certainclinicopathologic variables, including tumor stage (P ¼ 0.001),BMI (P ¼ 0.023), tumor size (P ¼ 0.001), lymph node (LN)metastasis (P¼ 0.003), and HER2 status (P¼ 0.013; Table 1). To

evaluate the association of clinicopathologic characteristics andserum visfatin levels as independent variables with patient sur-vival, univariate and multivariable Cox regression analyses wereemployed. In the univariate analysis, tumor size (e.g., P ¼ 0.002for >5 vs. <2 cm), LN metastasis (P < 0.001), hormone therapy(P ¼ 0.001), and serum visfatin (P ¼ 0.001) were significantlyassociated with OS (Table 2). In the adjusted multivariableanalysis, LN metastasis (P ¼ 0.009), hormone therapy (P ¼0.017), and serum visfatin (P ¼ 0.020) were significantly

Visfatin (ng/mL)0 12.5 25 50 100

Via

ble

cells

(% o

f con

trol)

80

100

120

140

160

No inhibitorImatinib (10 µmol/L)Stattic (1 µmol/L)Imatinib (10 µmol/L) + Stattic (1 µmol/L)

Low High

Ser

um v

isfa

tin(n

g/m

L)

20

30

40

50

p-STAT3 (Y705)

MDA-MB-231 MCF7 T-47DVi

able

cel

ls(%

of c

ontro

l)

80

100

120

140

160

0 ng/mL12.5 ng/mL25 ng/mL50 ng/mL100 ng/mL

Visf

atin

0 12.5 25 50 100 (ng/mL)Visfatin

β-Actin

MD

A-M

B-2

31

0 12.5 25 50 100 (ng/mL)

Visfatin

β-Actin

MC

F7

p-c-Abl (Y393/412)1.00 1.28 1.42 1.52 1.50

p-c-Abl (Y393/412)1.00 1.23 1.40 1.48 1.45

p-STAT3 (Y705)1.00 1.09 1.29 1.30 1.36

p-STAT3 (Y705)1.00 1.13 1.25 1.40 1.41

B

Score 0 Score 1p-STAT3 (Y705)

Score 2 Score 3

Score 0 Score 1

Score 2

p-c-Abl (Y393/412)

Score 3

C

D

E

A

p-STAT3 (S727)

Total STAT3

Total c-Abl

p-JAK2 (Y1007/1008)

Total JAK2

1.00 0.99 1.03 0.98 1.04

1.00 1.02 0.96 0.94 0.99

1.00 1.10 1.24 1.38 1.47

1.00 0.95 1.01 0.98 0.96

1.00 0.96 1.00 0.98 1.03

p-STAT3 (S727)

Total STAT3

Total c-Abl

p-JAK2 (Y1007/1008)

Total JAK2

1.00 0.98 0.96 0.97 1.01

1.00 1.12 1.20 1.31 1.42

1.00 0.94 0.98 1.00 0.95

1.00 0.97 0.96 0.99 1.02

1.00 1.04 0.99 1.02 1.03

p-c-Abl (Y393/412)Low High

Ser

um v

isfa

tin(n

g/m

L)

20

30

40

50

**

*******

******

******

***

*****

*****

****

*************

Figure 2.Effect of visfatin on breast cancer cellviability and signaling involving c-Abland STAT3. A, cell viability ofMDA-MB-231, MCF7, and T-47D cellstreated with visfatin (0–100 ng/mL)for 72 hours was measured by XTTcolorimetric assay; n ¼ 3. B, proteinexpression of c-Abl and STAT3signaling in MDA-MB-231 and MCF7cells treated with visfatin (0–100ng/mL) for 2 hours was analyzed byimmunoblotting. The data wererepresentative of three independentexperiments. C and D, proteinexpression of phospho-c-Abl (C) andphospho-STAT3 (D) in breast cancertissues was scored according tointensity of immunohistochemicalstaining (0, 1, 2, and 3). Low proteinexpression (scores 0 and 1; n ¼ 26 forphospho-c-Abl and n ¼ 47 forphospho-STAT3) or high proteinexpression (scores 2 and 3; n ¼ 74 forphospho-c-Abl and n ¼ 50 forphospho-STAT3) of phospho-c-Abl(C) and phospho-STAT3 (D) wasanalyzed for association of serumvisfatin levels from the correspondingbreast cancer patients; scale bar, 100mm. E, cell viability of MDA-MB-231cells pretreated with imatinib (0 or 10mmol/L) and Stattic (0 or 1 mmol/L) for1 hour followed by visfatin (0–100ng/mL) treatment for 72 hours wasmeasured by XTT colorimetric assay;n ¼ 3. The data were presented asmean � SD; � , P < 0.05; �� , P < 0.01;�� , P < 0.001; two-sided Studentt test.






Cel

l mig

ratio

n(%

of c

ontro

l)

0

50

100

150

200

250

Cel

l inv

asio

n(%

of c

ontro

l)

0

50

100

150

200

250

MM

P-9

sec

retio

n(%

of c

ontro

l)

0

50

100

150

200

MM

P-2

sec

retio

n(%

of c

ontro

l)

0

50

100

150

200

B

E

(–)

(+)

Inva

sion

MMP-2MMP-9

CVisfatin (100 ng/mL)

– + – +– – + +

D

MMP-2MMP-9

Visfatin (100 ng/mL)– + – +– – + +

MM

P-2

sec

retio

n(%

of c

ontro

l)

0

50

100

150

200

A– + – + – + – +– – + + – – + +– – – – + + + +

Imatinib (10 µmol/L)

Imatinib (10 µmol/L) Stattic (1 µmol/L)

Imatinib (10 µmol/L)Stattic (1 µmol/L) Stattic (1 µmol/L)

Visfatin (100 ng/mL)

48 h after microinjection

Visfatin No. of invasion (–)(% of total)

No. of invasion (+)(% of total) P

0 ng/mL 186(87.3%)

27(12.7%)

100 ng/mL 165(79.7%)

42(20.3%) 0.035

MM

P-9

sec

retio

n(%

of c

ontro

l)

0

50

100

150

200

******

***

– + – + – + – +– – + + – – + +– – – – + + + +Visfatin (100 ng/mL)

****

***

*** *** *

*** ** *** **

Figure 3.Effect of visfatin on breast cancer cell migration and invasion. A, migratory ability of MDA-MB-231 cells pretreated with imatinib (0 or 10 mmol/L) and Stattic(0 or 1 mmol/L) for 1 hour followed by visfatin (0 or 100 ng/mL) pretreatment for 24 hours was assessed by Transwell migration assay; n ¼ 3. B, invasive abilityof MDA-MB-231 cells pretreated with imatinib (0 or 10 mmol/L) and Stattic (0 or 1 mmol/L) for 1 hour followed by visfatin (0 or 100 ng/mL) pretreatment for24 hours was assessed by Transwell invasion assay; n ¼ 3; Scale bar, 20 mm. (Continued on the following page.)

Hung et al.





associated with OS (Table 2). Similar results were obtained fordisease-free survival on univariate and multivariable analyses(Supplementary Table S1).

We further studied the association of serum visfatin levels withadjuvant therapies for patient survival by Kaplan–Meier curveanalysis, and found that differential association only occurred inthe hormone therapy (HT) treatment group (P ¼ 0.027; Supple-mentary Fig. S1F). Serumvisfatin levels hadno effect onprognosisof OS in the other treatment groups (Supplementary Fig. S1A–S1E). Similar outcomes of the differential association of serumvisfatin levels with HT treatment were observed for disease-freesurvival (Supplementary Fig. S2) and disease recurrence (Supple-mentary Table S2).

Effect of visfatin on breast cancer cell viability and signalingNext, the biologic effects of visfatin on breast cancer cells were

explored both in vitro and in vivo. The assessment of cell viabilityby XTT assay showed a significant increase of viable cells in breastcancer cells treated with recombinant human visfatin (Pepro-Tech), includingMDA-MB-231 (e.g., P < 0.001 for visfatin 100 vs.0 ng/mL), MCF7 (e.g., P ¼ 0.001 for visfatin 100 vs. 0 ng/mL),T-47D (e.g., P < 0.001 for visfatin 100 vs. 0 ng/mL; Fig. 2A), andSKBR-3 cells (e.g., P ¼ 0.002 for visfatin 100 vs. 0 ng/mL;Supplementary Fig. S3A).

Activation of the cellular Abelson tyrosine kinase (c-Abl, alsoknown as ABL1) and signal transducer and activator of transcrip-tion 3 (STAT3) may act coordinately on tumor growth andmetastasis (26, 27). However, the roles of c-Abl and STAT3 inextracellular visfatin-mediated breast cancer have not beenreported before. We found that treatment of MDA-MB-231 cellswith visfatin resulted in increased levels of phosphorylated c-Ablat tyrosine-393/412 (Y393/412) residues (Fig. 2B). In addition,the levels of phosphorylated STAT3 at tyrosine-705 (Y705) res-idue andphosphorylated Janus kinase 2 (JAK2), a STAT3 activator(28), at tyrosine-1007/1008 (Y1007/1008) residues wereincreased in the cells treated with visfatin (Fig. 2B). Similar resultswere observed in MCF7 (Fig. 2B) and SKBR-3 cells (Supplemen-tary Fig. S3B).

We further investigated the association of phosphorylationlevels of c-Abl and STAT3 in breast tumor tissues with serumvisfatin levels from patients. The protein expression in tumortissues was examined by immunohistochemistry and scoredinto quartiles (0, 1, 2, and 3) as described previously (16). Thehigh expression group (scores 2 and 3; n ¼ 74) compared withthe low expression group (scores 0 and 1; n ¼ 26) of phos-phorylated c-Abl at Y393/412 in the tumor tissues was signif-icantly associated with increased serum visfatin levels deter-mined from the corresponding patients (P ¼ 0.002; Fig. 2C).Furthermore, the high expression group (scores 2 and 3; n ¼50) compared with the low expression group (scores 0 and 1;

n ¼ 47) of phosphorylated STAT3 at Y705 in the tumor tissueswas significantly associated with increased serum visfatin levelsfrom the corresponding patients (P < 0.001; Fig. 2D).

Although phosphorylation of STAT3may occur via activationof c-Abl in cells exposed to various cytokines and growth factors(26, 27), it remains unclear whether visfatin-mediated STAT3activation relies on c-Abl activity. We observed that the visfatin-induced phosphorylation of STAT3 at Y705 was decreased inthe presence of imatinib, an inhibitor of c-Abl kinase activityand clinical treatment of multiple cancers (Supplementary Fig.S3C; refs. 18, 29). Knockdown of the c-Abl protein expressionin breast cancer cells also resulted in reduced levels of phos-phorylated STAT3 at Y705 following visfatin treatment (Sup-plementary Fig. S3D). To further investigate the roles of c-Abland STAT3 activation in visfatin-mediated cell viability, MDA-MB-231 cells were treated with visfatin in combination withimatinib and Stattic, an inhibitor of STAT3 (19). The results ofXTT assay indicated that the visfatin-increased cell viability wassignificantly suppressed in cells pretreated with imatinib orStattic, or a combination of both (Fig. 2E).

Effect of visfatin on breast cancer cell migration andinvasion

In the in vitro Transwell migration assay, MDA-MB-231 cellsshowed a significant enhancement of migration after 24 hoursof visfatin pretreatment (P < 0.001), and the visfatin-inducedcell migration was significantly attenuated in the presence ofimatinib (P ¼ 0.002), Stattic (P ¼ 0.005), or combined ima-tinib and Stattic (P ¼ 0.006; Fig. 3A). The results of Matrigel-coated Transwell invasion assay indicated that MDA-MB-231cells pretreated with visfatin for 24 hours had a significantlyincreased capability of invasion compared with the controlcells (P < 0.001), which was reduced in the presence of imatinib(P ¼ 0.025), Stattic (P ¼ 0.016), or combined imatinib andStattic (P ¼ 0.004; Fig. 3B).

An important event in tumor metastasis is the degradationof extracellular matrix (ECM) and basement membranesthrough secreted matrix metalloproteinases (MMP), in partic-ular MMP-2 and MMP-9 (30). We examined the levels ofsecreted MMP-2 and MMP-9 by gelatin zymography and foundthat after 48 hours of visfatin treatment, the two MMPs in thesupernatant of MDA-MB-231 cells were significantly increasedcompared with control cells (P < 0.001 for MMP-2; P ¼ 0.001for MMP-9), and the visfatin-induced MMPs secretions weresignificantly decreased in the presence of imatinib (P ¼ 0.018for MMP-2; P ¼ 0.024 for MMP-9; Fig. 3C). In addition,the visfatin-induced MMPs secretions in the supernatant ofMDA-MB-231 cells were significantly reduced in the presenceof Stattic (P ¼ 0.002 for visfatin vs. Stattic þ visfatin in MMP-2;P ¼ 0.003 for visfatin vs. Stattic þ visfatin in MMP-9; Fig. 3D).

(Continued.) C, secretion of MMP-2 and MMP-9 in MDA-MB-231 cells pretreated with imatinib (0 or 10 mmol/L) for 1 hour followed by visfatin (0 or 100 ng/mL)treatment for 48 hours was assessed by gelatin zymographic analysis; n ¼ 3. D, secretion of MMP-2 and MMP-9 in MDA-MB-231 cells pretreated with Stattic(0 or 1 mmol/L) for 1 hour followed by visfatin (0 or 100 ng/mL) treatment for 48 hourswas assessed by gelatin zymographic analysis; n¼ 3. The datawere presentedasmean� SD; � , P <0.05; �� , P <0.01; �� , P <0.001; two-sided Student t test. E, in vivo invasive ability of MDA-MB-231 cells pretreatedwith visfatin (0 or 100 ng/mL)for 24 hours was assessed in xenograft zebrafish model. After 48 hours of microinjection with DiI-labeled MDA-MB-231 cells into the perivitelline cavity ofzebrafish, distribution of the MDA-MB-231 cells in whole zebrafishwas inspected under fluorescent microscope. Micrographswere representatives for zebrafishwithno invasive (top) or with invasive (bottom) MDA-MB-231 cells, and arrowheads in the lower micrograph indicated invasive MDA-MB-231 cells observed at thelocation other than the original microinjection site in the perivitelline cavity. The inset was enlarged from a local view. A P value was determined by two-sided c2 test.n ¼ 213 for the group with 0 ng/mL of visfatin and n ¼ 207 for the group with 100 ng/mL of visfatin. Red, DiI-labeled MDA-MB-231 cells. Green, enhanced GFP(EGFP)-expressed vessels of zebrafish; scale bar, 200 mm.






We further evaluated whether the in vivo invasiveness ofbreast cancer cells could be altered by visfatin in a xenograftzebrafish model (23). As shown in Fig. 3E, MDA-MB-231cells pretreated with visfatin for 24 hours before microin-jection of the cells into the perivitelline cavity of zebrafishshowed a significantly increased degree of cell invasioncompared with the control cells without visfatin pretreat-ment (P ¼ 0.035).

Effect of visfatin on anchorage-independent growth in vitroThe effect of visfatin on in vitro anchorage-independent

growth was examined by soft agar colony formation assay.MDA-MB-231 cells treated with visfatin showed significantlyincreased colony formation compared with the control cellswithout visfatin treatment (e.g., P ¼ 0.005 for visfatin 100 vs.0 ng/mL; Fig. 4A). Although less capable of forming anchor-age-independent growth than MDA-MB-231 cells (31), MCF7cells treated with visfatin also showed significantly increasedcolony formation compared with the control cells withoutvisfatin treatment (P ¼ 0.002 for visfatin 100 vs. 0 ng/mL;Fig. 4A).

Anoikis, a specific form of cell death induced by cell detach-ment fromECM, is one of themechanisms for avoidingmetastaticcolonization via prevention of anchorage-independent growth ofcancer cells (32). We found that MDA-MB-231 cells treated withvisfatin revealed a significant reduction of anoikis cell death (P¼0.006; Supplementary Fig. S4A).

Effect of visfatin on tumor growth and metastasis in miceTo further confirm the association of visfatin with tumor

formation and metastasis in vivo, we used an orthotopic xeno-graft model in nude mice, in which MDA-MB-231 cells wereinjected into the mammary fat pads of female mice, followedby intraperitoneal administration of visfatin (0.5 mg/kg) ornormal saline as vehicle control three times a week. As shownin Fig. 4B, the mice administered with visfatin showed asignificant increase in orthotopic tumor growth compared withthe control mice (P ¼ 0.042 for tumor volume at week 8). Allmice were sacrificed after 8 weeks of treatment, and the weightof orthotopic tumors from the mice administered with visfatinwas found to be significantly higher than those from the controlmice (P ¼ 0.003; Fig. 4C).

The phosphorylated levels of c-Abl and STAT3 in the ortho-topic tumors were examined after the mice were sacrificed.Immunohistochemical analysis revealed that the expression ofphosphorylated c-Abl at Y393/412 in the orthotopic tumorsfrom visfatin-treated mice was significantly higher than thatfrom the control mice (P ¼ 0.002; Fig. 4D). There was also asignificant increase of phosphorylated STAT3 at Y705 in theorthotopic tumors from the visfatin-treated mice comparedwith the control mice (P ¼ 0.011; Fig. 4E). We further inves-tigated for the occurrence of distant lung metastasis by immu-nohistochemical analysis. Two tumor marker proteinsexpressed in MDA-MB-231 cells, cytokeratin 18 (CK18) andvimentin (24), were significantly increased in the lung tissuesof mice treated with visfatin compared with those from thecontrol mice (P < 0.001 for CK18; P ¼ 0.009 for vimentin; Fig.4F and G, respectively). The body weight of the mice during thecourse of treatment was not significantly different in eachgroup (Supplementary Fig. S5A), neither were the levels ofalanine aminotransferase (ALT), aspartate aminotransferase

(AST), blood urea nitrogen (BUN), and creatinine for liverand renal functions of the mice (Supplementary Fig. S5B).

The role of c-Abl activation in visfatin-mediated metastasiswas further studied by in vivo bioluminescent imaging using theintravenous tail vein injection model in nude mice. The cellviability of luciferase-expressing MDA-MB-231 (MDA-MB-231-Luc) cells was not significantly different in each group beforethe tail vein injection (Supplementary Fig. S4B). After 7 dayspost-injection, the mice injected with the visfatin-pretreatedMDA-MB-231-Luc cells revealed a significantly higher biolu-minescent signal in lung tissues than the mice injected withcontrol MDA-MB-231-Luc cells (P ¼ 0.048), whereas the miceinjected with the cells pretreated with imatinib followed byvisfatin pretreatment had a significantly reduced biolumines-cent signal in lung tissues (P ¼ 0.046; Fig. 4H).

DiscussionSerum visfatin as a biomarker for clinical evaluation in breastcancer

Serumvisfatin has previously been shown to be associatedwithbreast cancer progression (7–11). Yet, its prognostic valueremains insufficiently assessed, except for Li and colleagues(10) who reported that high serum visfatin level was associatedwithpoorer survival of breast cancer patients. Although a similarlyunfavorable outcome for the survival of the patients with highlevel of serum visfatin was observed in the present study, our datafurther analyzed important clinical factors that have not beenpreviously elucidated. For instance, we reported that the patientswith both high serum visfatin level and ER-negative status had theworst survival compared with the other groups of patients. More-over, the ER-negative patients with high serum visfatin levelwithout receiving hormone therapy (HT�) had a poorer survivaland more recurrent incidence than the ER-positive patients withhigh serum visfatin level receiving hormone therapy (HTþ),whereas there was no significant difference between HT� andHTþ treatment groups of the patients with low serum visfatin.Together, the data suggest that different serum visfatin levels mayhave clinical relevance as a prognostic tool in breast cancer.Although the mechanisms underlying the relationship betweenserumvisfatin levels and the patient response to hormone therapyare not clear at now, we propose that it may be linked through anindirect regulation in adipocytes, whose ER activation can lead toupregulation of visfatin expression (33, 34). As adipocytes havebeen known the major source to release visfatin (35, 36), thepatients with anti-ER treatment may cause the reduction ofvisfatin release from adipocytes into circulation, which in turnalleviates accumulationof serumvisfatin thatmaypromote breastcancer progression. However, more experiments will be requiredto examine this hypothesis.

In this study, the level of serum visfatin in breast cancer patients(40.87 � 13.86 ng/mL) was observed to be higher than normalfemale participants (32.20 � 17.42 ng/mL). Analysis of clinico-pathologic parameters in our study revealed that high level ofserum visfatin (cutoff at 33.75 ng/mL) was associated with tumorstage, BMI, tumor size, LNmetastasis, and HER2 status (Table 1).Among these clinicopathologic parameters, tumor stage, and LNmetastasis were commonly found to be associated with higherserumvisfatin level in this study and others (7, 9, 11), suggesting apotentially unfavorable role of serum visfatin in breast cancer. Itwas noted that in our clinicopathologic analyses, status of HER2

Hung et al.





Control Visfatin

Tum

or w

eigh

t(g

)

0.00.20.40.60.81.01.21.4

Visfatin (ng/mL)0 50 100

Col

ony

form

atio

n(%

of c

ontro

l)

0

100

200

300

400 MDA-MB-231

Control Visfatin

Vim

entin

Exp

ress

ion

(H-s

core

)

0

50

100

150

200

Control Visfatin

CK

18 E

xpre

ssio

n(H

-sco

re)

0

50

100

150

200

250

A

F

0 50 100

Visfatin (ng/mL)

MDA-MB-231

MCF7

CControl

Visfatin

Control Visfatin G

Week1 2 3 4 5 6 7 8

Tum

or v

olum

e(m

m3 )

0

250

500

750

1,000

1,250

1,500 ControlVisfatin

B

*

D Control Visfatin E

Control Visfatin

Control Visfatinp-c-

Abl

(Y39

3/41

2) E

xpre

ssio

n(H

-sco

re)

0

50

100

150

200

250

Control Visfatinp-S

TAT3

(Y70

5) E

xpre

ssio

n(H

-sco

re)

0

50

100

150

Control Visfatin

*

**

Visfatin (ng/mL)0 50 100

Col

ony

form

atio

n(%

of c

ontro

l)

0

100

200

300 MCF7

**

**

** *

*****

MDA-MB-231

ControlImatinib

Visfatin

Imatinib+Visfatin

Lum

ines

cenc

e(p

hoto

n/se

cond

/cm

2 /sr)

0

50 × 103

100 × 103

150 × 103

200 × 103HImatinibVisfatin

– + – +– – + +

Luminescence* *2.5

2.0

1.5

1.0

(x105)

MDA-MB-231-Luc

Figure 4.Effect of visfatin on breast tumorgrowth and metastasis. A, in vitroanchorage-independent growth ofMDA-MB-231 and MCF7 cells treatedwith visfatin (0, 50, and 100 ng/mL)wasdetermined by soft agar colonyformation assay after 3weeks in culture;n ¼ 3; scale bar, 200 mm. B, in vivobreast tumor growth was assessed bythe orthotopic xenograft model in nudemice injected with MDA-MB-231 cells tothe 4th mammary fat pads. The micewere intraperitoneally administeredvisfatin (0.5 mg/kg) or normal saline ascontrol three times a week (n ¼ 10/group). Tumor volumes were measuredweekly and calculated with the formulaof (width2 � length)/2 for 8 weeks. C,after 8 weeks as described in B, all micewere sacrificed and the orthotopictumors were collected for tumor weightmeasurement. The inset showed fiverepresentative orthotopic tumors foreach group dissected from the mice;scale bar, 10 mm. D to G, after sacrificeof the mice, the orthotopic tumors werecollected for protein expression ofphospho-c-Abl (D) and phospho-STAT3 (E), and the lung tissues werecollected for metastatic assessment forprotein expression of cytokeratin 18(CK18; F) and vimentin (G) byimmunohistochemical analysis withH-score; scale bar, 100 mm. H, in vivometastasis of luciferase-expressingMDA-MB-231 (MDA-MB-231-Luc)cells pretreated with imatinib (0 or10 mmol/L) for 1 hour followed byimatinib (0 or 10 mmol/L) combinedvisfatin (0 or 100 ng/mL) pretreatmentfor 24 hours was intravenously injectedinto nude mice via the lateral tail vein(n ¼ 3–4 per group). The micecontaining MDA-MB-231-Luc cellspretreated with imatinib alone,or with imatinib followed by visfatinpretreatment, were intraperitoneallyinjected with imatinib (50 mg/kg daily).After 7 days of tail vein injection, themice were anesthetized andintraperitoneally injected with D-luciferin (150 mg/kg) for detection ofbioluminescence by IVIS Spectrum invivo imaging system. The data werepresented as mean � SD; � , P < 0.05;�� , P < 0.01; �� , P < 0.001; two-sidedStudent t test.






protein expression in breast tumor tissues was positively corre-latedwith serum visfatin level in patients.We further analyzed theclinical data with classification of molecular subtypes in breastcancer (i.e., luminal A, luminal B, HER2, and triple-negative;ref. 37), and the results consistently suggested that HER2 expres-sion might be associated with high serum visfatin (P ¼ 0.018 forluminal A vs. luminal B; Supplementary Fig. S6A). We alsoexamined the effect of visfatin on HER2 expression in culturedbreast cancer cells, and found that visfatin treatment resulted in anincreased level of HER2 protein expression by immunoblotting(Supplementary Fig. S6B). Database searching through the use ofIngenuity Pathway Analysis (IPA; QIAGEN) revealed that JAK2,STAT3, NF-kB, IL6, and TGFb formed a potential regulatorynetwork between visfatin and HER2 in several cell types (Sup-plementary Fig. S6C). As none of these molecules were identifiedin breast cancer cells, it will be worthwhile to evaluate the cellularsignaling, including that if any of thesemoleculesmayhave a role,in visfatin-mediated HER2 expression in breast cancer cells.

Characterization of the biologic effect of visfatin on breastcancer cells

Our in vitro data showed that extracellular visfatin promotedbreast cancer cell proliferation, migration, invasion, andMMP2/9secretion. In addition, our in vivo data using zebrafish and nudemice as xenograft models further confirmed the promoting effectof extracellular visfatin on breast tumor growth and metastasis.Notably, we provided the first evidence for the involvement ofc-Abl and STAT3, two important oncoproteins (38–40), in extra-cellular visfatin-promoted breast cancer. Although the intracellu-lar reactive oxygen species (ROS)was reportedly to beupregulatedby visfatin stimulation in some cell types (41–43), it was notdetectable in our current treatment with visfatin in breast cancercells (Supplementary Fig. S3E and S3F). Other mechanismsreported in extracellular visfatin–mediated breast cancer growthandmetastatic potential included cyclinD1 and cdk2 (44), NF-kBand Notch1 (45), or PI3K/Akt and TGFb (46). As the identity of"visfatin receptor" is not known to date, it limits our understand-ing for the molecular mechanisms of extracellular visfatin. Nev-ertheless, our current results raised a possibility that the putativereceptor for visfatin could belong to the receptor family thattransduces signals via c-Abl and STAT3, such as EGF, platelet-derived growth factor (PDGF), insulin receptor, or insulin-likegrowth factor (IGF) receptors (Supplementary Fig. S6D; refs. 26,38, 39, 47). Whether there is a direct interaction between c-Abland the putative receptor activated by visfatin, or a direct inter-action between c-Abl and STAT3 after visfatin stimulation,remains to be determined.

Stratifying patients according to serum visfatin levels forpotential targeted adjuvant therapy

Previous clinical trials have shown no clinical benefit fromimatinib in breast cancer patients (48, 49). However, this may bedue to a selection issue, as patients have been comprehensivelyincluded without stratifying those that may benefit most fromimatinib treatment. Until now, there has been no clinical markerthat may allow such stratification. Our animal studies showedthat the lung metastasis of breast cancer cells pretreated withvisfatin via tail vein xenograft in nudemicewas largely suppressedby imatinib. Equally as important, there was no significant effecton metastasis with imatinib treatment in the group withoutexposure to visfatin, further providing a strong argument for

stratification of treatment with imatinib according to extracellularlevels of visfatin. The potential clinical importance of a targetedanti-metastatic agent in breast cancer patients, using an alreadywidely available but unused drug in breast cancer, cannot beunderestimated.Wewould suggest that future clinical trials on theefficacy of imatinib may stratify patients according to visfatinlevels and select those with pre-metastatic, high serum visfatinbreast cancer. In addition, this may advantageously offer alter-native therapeutic options in the triple-negative breast cancerpatients.

Study limitationsThere are a few limitations to this study, including the lack of a

standard definition of high or low serum visfatin level and sizeand geographic localization of study population. Therefore, caremust be taken in interpreting clinical analyses from this study andothers. In addition, although we showed the effects of extracel-lular visfatin on breast cancer cells in both in vitro and in vivoexperiments that mainly used MDA-MB-231 cells for metastasisand animal studies, it may not fully represent heterogeneoushuman breast cancer behaviors in clinical settings. It should alsobe noted that imatinib is known to be a tyrosine kinase inhibitorfor c-Abl, c-Kit, PDGFR, and DDR (50).

ConclusionsIn conclusion, our clinical data suggested that high level of

serumvisfatinwas associatedwithmalignant breast cancer behav-ior, and the level of serum visfatin could offer a means forprognosis in breast cancer. Moreover, the results of our in vitroand in vivo experiments revealed that extracellular visfatin pro-moted breast cancer cell growth and metastatic ability via acti-vation of c-Abl and STAT3. There are several key implicationsresulting from this study. First, stratification of breast cancerpatients by serum visfatin levels may help identify patients thatwould benefit from inhibition of downstream effectors of visfatinsignaling. Second, categorization of the patients according to theER status may provide information for risk assessment of thesurvival of patients with different levels of serum visfatin. Third,our study is the first to suggest that the use of imatinib may havetherapeutic potential in the treatment of breast cancer for thegroup of patients with higher levels of serum visfatin.

Further studies, including discovery of the receptor for visfatin,development of specific inhibitors for extracellular visfatin, andclinical trials stratifying patients for treatment according to dif-ferent serum visfatin levels, will be potentially important to thefuture design of targeted therapies in consideration of differentserum visfatin levels in breast cancer.

Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.

Authors' ContributionsConception and design: A.C. Hung, C.-H. Tsai, Y.-M. Wang, S.-S.F. YuanDevelopment of methodology: A.C. Hung, S. Lo, C.-H. Tsai, W. Liu, Y.-H. Lo,S.-S.F. YuanAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): A.C. Hung, M.-F. Hou, Y.-C. Lee, C.-H. Tsai,Y.-Y. Chen, W. Liu, Y.-H. SuAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): A.C. Hung, S. Lo, Y.-C. Lee, C.-H. Tsai, Y.-H. Lo,S.-C. Wu, S.C.-S. Hu, S.-S.F. Yuan


Hung et al.




Writing, review, and/or revision of the manuscript: A.C. Hung, S. Lo,Y.-C. Hsieh, S.C.-S. Hu, M.-H. Tai, Y.-M. Wang, S.-S.F. YuanAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): A.C. Hung, C.-H. Tsai, W. Liu, Y.-H. Lo, C.-H.Wang, S.-S.F. YuanStudy supervision: M.-F. Hou, Y.-M. Wang

Grant SupportThis work was supported by grants from Kaohsiung Medical University

Hospital (KMUH102-2T07 and KMUHI102-2R25), Kaohsiung Medical Uni-

versity (aim for the top journals grant, KMU-DT103010 and KMU-TP103D18),National Health Research Institutes (NHRI-EX100-9829BI and NHRI-EX104-10212BI), and Ministry of Health and Welfare (MOHW103-TD-B-111-05) ofTaiwan.

The costs of publication of this articlewere defrayed inpart by the payment ofpage charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received November 6, 2015; revised March 14, 2016; accepted March 22,2016; published OnlineFirst April 1, 2016.

References1. Vona-Davis L, Rose DP. Adipokines as endocrine, paracrine, and autocrine

factors in breast cancer risk and progression. Endocr Relat Cancer2007;14:189–206.

2. Vansaun MN. Molecular pathways: adiponectin and leptin signaling incancer. Clin Cancer Res 2013;19:1926–32.

3. Samal B, Sun Y, Stearns G, Xie C, Suggs S, McNiece I. Cloning andcharacterization of the cDNA encoding a novel human pre–B-cell colo-ny-enhancing factor. Mol Cell Biol 1994;14:1431–7.

4. Dahl TB, Holm S, Aukrust P, Halvorsen B. Visfatin/NAMPT: a multifacetedmolecule with diverse roles in physiology and pathophysiology. Annu RevNutr 2012;32:229–43.

5. Shackelford RE, Mayhall K, Maxwell NM, Kandil E, Coppola D. Nicotin-amide phosphoribosyltransferase in malignancy: a review. Genes Cancer2013;4:447–56.

6. Lee YC, Yang YH, Su JH, Chang HL, Hou MF, Yuan SS. High visfatinexpression in breast cancer tissue is associated with poor survival. CancerEpidemiol Biomarkers Prev 2011;20:1892–901.

7. DalamagaM, Archondakis S, SotiropoulosG, Karmaniolas K, PelekanosN,Papadavid E, et al. Could serum visfatin be a potential biomarker forpostmenopausal breast cancer? Maturitas 2012;71:301–8.

8. Dalamaga M, Karmaniolas K, Papadavid E, Pelekanos N, Sotiropoulos G,Lekka A. Elevated serum visfatin/nicotinamide phosphoribosyl-transferaselevels are associated with risk of postmenopausal breast cancer indepen-dently from adiponectin, leptin, and anthropometric and metabolic para-meters. Menopause 2011;18:1198–204.

9. Assiri AM, Kamel HF, Hassanien MF. Resistin, visfatin, adiponectin, andleptin: risk of breast cancer in pre- and postmenopausal saudi females andtheir possible diagnostic and predictive implications as novel biomarkers.Dis Markers 2015;2015:253519.

10. Li XY, Tang SH, Zhou XC, Ye YH, Xu XQ, Li RZ. Preoperative serum visfatinlevels and prognosis of breast cancer among Chinese women. Peptides2014;51:86–90.

11. Assiri AM, Kamel HF. Evaluation of diagnostic and predictive value ofserum adipokines: leptin, resistin and visfatin in postmenopausal breastcancer. Obes Res Clin Pract 2015;S1871–403X:00131–3.

12. Carlson RW, Anderson BO, Bensinger W, Cox CE, Davidson NE, Edge SB,et al. NCCN practice guidelines for breast cancer. Oncology 2000;14:33–49.

13. The World Health Organization. Histological typing of breast tumors.Neoplasma 1983;30:113–23.

14. FriersonHF Jr, Wolber RA, Berean KW, Franquemont DW,GaffeyMJ, BoydJC, et al. Interobserver reproducibility of the Nottingham modification ofthe Bloomand Richardson histologic grading scheme for infiltrating ductalcarcinoma. Am J Clin Pathol 1995;103:195–8.

15. Singletary SE, Allred C, Ashley P, Bassett LW, Berry D, Bland KI, et al.Revision of the American Joint Committee on Cancer staging system forbreast cancer. J Clin Oncol 2002;20:3628–36.

16. Sato T, Neilson LM, Peck AR, Liu C, Tran TH, Witkiewicz A, et al. Signaltransducer and activator of transcription-3 andbreast cancer prognosis. AmJ Cancer Res 2011;1:347–55.

17. Stoyianni A, Goussia A, Pentheroudakis G, Siozopoulou V, Ioachim E,Krikelis D, et al. Immunohistochemical study of the epithelial–mesenchy-mal transition phenotype in cancer of unknown primary: incidence,correlations and prognostic utility. Anticancer Res 2012;32:1273–81.

18. Baker SJ, Reddy EP. Targeted inhibition of kinases in cancer therapy. MtSinai J Med 2010;77:573–86.

19. Schust J, Sperl B, Hollis A, Mayer TU, Berg T. Stattic: a small-moleculeinhibitor of STAT3 activation and dimerization. Chem Biol 2006;13:1235–42.

20. Chen YJ, Hung CM, Kay N, Chen CC, Kao YH, Yuan SS. Progesteronereceptor is involved in 2,3,7,8-tetrachlorodibenzo-p-dioxin-stimulatedbreast cancer cells proliferation. Cancer Lett 2012;319:223–31.

21. Yuan SS, Hou MF, Hsieh YC, Huang CY, Lee YC, Chen YJ, et al. Role ofMRE11 in cell proliferation, tumor invasion, and DNA repair in breastcancer. J Natl Cancer Inst 2012;104:1485–502.

22. Westerfield M. The zebrafish book: a guide for the laboratory use ofzebrafish (Brachydanio rerio). University of Oregon Press, Eugene;1993, 2nd edition. p. 300.

23. Lee SL, Rouhi P, Dahl Jensen L, Zhang D, Ji H, Hauptmann G, et al.Hypoxia-induced pathological angiogenesis mediates tumor cell dissem-ination, invasion, and metastasis in a zebrafish tumor model. Proc NatlAcad Sci U S A 2009;106:19485–90.

24. Tiang JM, Butcher NJ, Minchin RF. Effects of human arylamine N-acetyl-transferase I knockdown in triple-negative breast cancer cell lines. CancerMed 2015;4:565–74.

25. Schito L, Rey S, Tafani M, Zhang H, Wong CC, Russo A, et al. Hypoxia-inducible factor 1-dependent expression of platelet-derived growth factor Bpromotes lymphatic metastasis of hypoxic breast cancer cells. Proc NatlAcad Sci U S A 2012;109:E2707–16.

26. Greuber EK, Smith-Pearson P, Wang J, Pendergast AM. Role of ABL familykinases in cancer: from leukaemia to solid tumours. Nat Rev Cancer2013;13:559–71.

27. Buettner R, Mora LB, Jove R. Activated STAT signaling in human tumorsprovides novel molecular targets for therapeutic intervention. Clin CancerRes 2002;8:945–54.

28. Quintas-Cardama A, Verstovsek S. Molecular pathways: Jak/STAT path-way: mutations, inhibitors, and resistance. Clin Cancer Res 2013;19:1933–40.

29. Gold JS, van der Zwan SM, GonenM,Maki RG, Singer S, BrennanMF, et al.Outcome of metastatic GIST in the era before tyrosine kinase inhibitors.Ann Surg Oncol 2007;14:134–42.

30. John A, Tuszynski G. The role of matrix metalloproteinases in tumorangiogenesis and tumor metastasis. Pathol Oncol Res 2001;7:14–23.

31. Uehara N, Matsuoka Y, Tsubura A. Mesothelin promotes anchorage-independent growth and prevents anoikis via extracellular signal-regulatedkinase signaling pathway in human breast cancer cells. Mol Cancer Res2008;6:186–93.

32. Paoli P, Giannoni E, Chiarugi P. Anoikismolecular pathways and its role incancer progression. Biochim Biophys Acta 2013;1833:3481–98.

33. MacLaren R, Cui W, Cianflone K. Visfatin expression is hormonallyregulated by metabolic and sex hormones in 3T3-L1 pre-adipocytes andadipocytes. Diabetes Obes Metab 2007;9:490–7.

34. Zhou J, Seidel ER. Estrogens induce visfatin expression in 3T3-L1 cells.Peptides 2010;31:271–4.

35. Fain JN, Tagele BM, Cheema P, Madan AK, Tichansky DS. Release of 12adipokines by adipose tissue, nonfat cells, and fat cells from obese women.Obesity 2010;18:890–6.

36. Tanaka M, Nozaki M, Fukuhara A, Segawa K, Aoki N, Matsuda M, et al.Visfatin is released from 3T3-L1 adipocytes via a non-classical pathway.Biochem Biophys Res Commun 2007;359:194–201.

37. Holliday DL, Speirs V. Choosing the right cell line for breast cancerresearch. Breast Cancer Res 2011;13:215.






38. Ganguly SS, Plattner R. Activation of abl family kinases in solid tumors.Genes Cancer 2012;3:414–25.

39. Srinivasan D, Sims JT, Plattner R. Aggressive breast cancer cells are depen-dent on activated Abl kinases for proliferation, anchorage-independentgrowth and survival. Oncogene 2008;27:1095–105.

40. Yu H, Kortylewski M, Pardoll D. Crosstalk between cancer and immunecells: role of STAT3 in the tumour microenvironment. Nat Rev Immunol2007;7:41–51.

41. KimSR, BaeYH, Bae SK,Choi KS, YoonKH,KooTH, et al. Visfatin enhancesICAM-1 and VCAM-1 expression through ROS-dependent NF-kappaBactivation in endothelial cells. Biochim Biophys Acta 2008;1783:886–95.

42. Oita RC, Ferdinando D, Wilson S, Bunce C, Mazzatti DJ. Visfatininduces oxidative stress in differentiated C2C12 myotubes in an Akt-and MAPK-independent, NFkB-dependent manner. Pflugers Arch2010;459:619–30.

43. Song SY, Jung EC, Bae CH, Choi YS, Kim YD. Visfatin induces MUC8 andMUC5B expression via p38 MAPK/ROS/NF-kappaB in human airwayepithelial cells. J Biomed Sci 2014;21:49.

44. Kim JG, Kim EO, Jeong BR, Min YJ, Park JW, Kim ES, et al. Visfatinstimulates proliferation of MCF-7 human breast cancer cells. Mol Cells2010;30:341–5.

45. ParkHJ, Kim SR, Kim SS,WeeHJ, BaeMK, RyuMH, et al. Visfatin promotescell and tumor growth by upregulatingNotch1 in breast cancer.Oncotarget2014;5:5087–99.

46. Soncini D, Caffa I, Zoppoli G, Cea M, Cagnetta A, Passalacqua M, et al.Nicotinamide phosphoribosyltransferase promotes epithelial-to-mesen-chymal transition as a soluble factor independent of its enzymatic activity.J Biol Chem 2014;289:34189–204.

47. Peng Q, Jia SH, Parodo J, Ai Y, Marshall JC. Pre-B cell colony enhancingfactor induces Nampt-dependent translocation of the insulin receptor outof lipid microdomains in A549 lung epithelial cells. Am J Physiol Endo-crinol Metab 2015;308:E324–33.

48. Cristofanilli M, Morandi P, Krishnamurthy S, Reuben JM, Lee BN, FrancisD, et al. Imatinib mesylate (Gleevec) in advanced breast cancer-expressingC-Kit or PDGFR-beta: clinical activity and biological correlations. AnnOncol 2008;19:1713–9.

49. Modi S, Seidman AD, Dickler M, Moasser M, D'Andrea G, Moynahan ME,et al. A phase II trial of imatinib mesylate monotherapy in patients withmetastatic breast cancer. Breast Cancer Res Treat 2005;90:157–63.

50. Bantscheff M, Eberhard D, Abraham Y, Bastuck S, Boesche M, Hobson S,et al. Quantitative chemical proteomics reveals mechanisms of action ofclinical ABL kinase inhibitors. Nat Biotechnol 2007;25:1035–44.


Hung et al.




2016;22:4478-4490. Published OnlineFirst April 1, 2016.Clin Cancer Res Amos C. Hung, Steven Lo, Ming-Feng Hou, et al. Cancer Is Mediated Through c-Abl and STAT3 ActivationExtracellular Visfatin-Promoted Malignant Behavior in Breast

Updated version

10.1158/1078-0432.CCR-15-2704doi:

Access the most recent version of this article at:

Material

Supplementary

http://clincancerres.aacrjournals.org/content/suppl/2016/04/01/1078-0432.CCR-15-2704.DC1

Access the most recent supplemental material at:

Cited articles

http://clincancerres.aacrjournals.org/content/22/17/4478.full#ref-list-1

This article cites 48 articles, 12 of which you can access for free at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected]

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://clincancerres.aacrjournals.org/content/22/17/4478To request permission to re-use all or part of this article, use this link



http://clincancerres.aacrjournals.org/lookup/doi/10.1158/1078-0432.CCR-15-2704

http://clincancerres.aacrjournals.org/content/suppl/2016/04/01/1078-0432.CCR-15-2704.DC1

http://clincancerres.aacrjournals.org/content/22/17/4478.full#ref-list-1

http://clincancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://clincancerres.aacrjournals.org/content/22/17/4478


Extracellular Visfatin-Promoted Malignant Behavior in ... · Biology of Human Tumors Extracellular...

Documents

Transcript of Extracellular Visfatin-Promoted Malignant Behavior in ... · Biology of Human Tumors Extracellular...